CONSTRUCTIVE INDUCTION APPROACH TO MACHINE LEARNING BASED ON FUNCTIONAL DECOMPOSITION OF RELATIONS

Funding

This is a joint effort with Professor Malgorzata Marek-Sadowska from University of California Santa-Barbara, Professor Lech Jozwiak from Technical University of Eindhoven, the Netherlands, and Professor Tadeusz Luba from Warsaw University of Technology, Poland.

Ashenhurst/Curtis Decomposition has been extended to Multi-Valued Functions and Relations. We are able to decompose large Machine Learning benchmarks such as breastc or different variants of trains .

Papers:

Talks:

Slides of presentation of Craig Files at ISMVL'97:
Decomposition of Multi-valued Functions.

Applications:

Medical Information for ML Projects

Information on Ovulation Research and Products

Reconfigurable Computing Commercial Info.